Professional Documents
Culture Documents
The Design and Implementation of a Database System
The Design and Implementation of a Database System
The Design and Implementation of a Database System
net/publication/228957419
CITATIONS READS
0 7,383
2 authors:
All content following this page was uploaded by Schubert Foo on 20 May 2014.
Schubert Foo
Division of Information Studies, School of Applied Science,
Nanyang Technological University, Singapore
Email: assfoo@ntu.edu.sg Telephone: (65) 790-4621 Fax: (65) 792-6559
Mark Hepworth
Department of Information Science
Loughborough University, United Kingdom
Email: M.Hepworth@lboro.ac.uk
Abstract
One of the most difficult aspects of deriving an information management plan for a
large knowledge based enterprise is to determine what information people require to
do their various tasks as well as help them expand and update their existing
knowledge. Defining the information needs of the organisation as a whole is a
necessary part of a systematic process of becoming such an enterprise.
This paper provides an overview of a methodology used to elicit the information
needs of an organisation. The focus however is on the design and implementation of
two databases that were used to support this process. These databases proved to be
extremely useful for organising the qualitative and quantitative data derived from the
information needs analysis research and also for analysing and presenting the
findings. The organization, namely, the Singapore Productivity and Standards Board,
whose role is to enhance the productivity of business and industry in Singapore,
employs approximately 1000 people. The use of database technology combined with
networked access to staff facilitated the capture and analysis of the information needs
of a large number of people in a short period of time. A description is given of the
database design; database implementation; the queries supported by the database as
well as the exporting of data to other applications to help visualize information needs.
1. Introduction
The Singapore Productivity and Standards Board (PSB) was formed in 1996 through
the merger of the National Productivity Board and the Singapore Institute of
Standards and Industrial Research. The merger was designed to enable an integrated
approach to the management of Total Factory Productivity to sustain Singapore’s
future growth (PSB, 1997). In order to sustain this growth, Singapore has seen a
gradual change in emphasis from a low-cost manufacturing economy towards a higher
1
value-added knowledge based economy (or information society) that emphasizes the
importance of human and intellectual capital for the future.
Taking the lead the PSB has recently embarked upon an information architecture
project. The term information architecture is used here to mean a description of the
information that people need and use across the organisation, where it comes, how
they use it and what happens to it. The availablity and perceived value of information
categories are also addressed. The resulting information architecture is therefore a
'map' of the organisation's information environment on the basis of which information
policy including data acquisition, data storage and IT solutions can be planned. The
initiative to undertake this exercise stems from the realisation that systematic access,
management, use and exploitation of data, information and knowledge can have
significant bearing on an organization’s competitiveness (Davenport & Marchand,
1999; Davenport, 1998). Knowledge-based enterprises are beginning to emerge.
Organisations such as Monsanto and Glaxo Welcome are quoted as being or are
becoming such organisations, (Syrme & Amidon, 1997).
In this paper, the authors, who acted as consultants to the project, trace the design and
development of an integrated set of information needs databases that were used to
assist in the derivation of the information architecture. The focus of the article is on
the design and usefulness of the databases that facilitated the management of project
data. A separate article is planned to discuss in detail the combination of qualitative
and quantitative methods used to derive the data. However an outline of the approach
taken is given below.
2
scale of the operation and the time constraints (three months) specific techniques had
to be developed. These techniques included the use of both qualitative and
quantitative techniques to understand the needs of the people in the organisation.
These techniques were based on a theoretical framework derived form user studies,
human computer interface design and software engineering. Similarly to Orna (1999)
emphasis was placed on understanding what information helped achieve
organisational goals. Emphasis however was placed on task analysis as a means for
eliciting information needs. Selective sampling and the use of specific techniques
such as flowcharting during focus groups helped to get input from a diverse range of
staff and to incrementally build a picture of the needs of respondents. This was
facilitated to a great extent by the use of information technology including database
technology, networked access to respondents and the use of applications to help
visualise the findings. Networked access in particular enabled the testing results of
the qualitative research on the majority of people in the organisation. It also enabled
the gathering of quantitative data on people's assessment of the availability and value
of categories. Increasingly software is being used to help visualise the relationships
between data. Horton (1994) was one of the first researchers to consciously use
database techniques to understand information use in the organisation. However
Horton was primarily concerned with mapping the information that exists in the
organisation i.e. an information audit rather than identifying the fundamental
information needs of people in an organisation.
2. Methodology
PSB is a large organization that employs approximately 1000 staff the majority of
whom are professional graduate staff. It encompasses 13 Divisions and comprises in
excess of 50 departments, centres and strategic business units. More than 20 separate
information systems exist to support various functions for the organization. Due to
the large number of staff and the three-month time constraint placed on the project
data collection had to be particularly effective. The developed approach had to satisfy
a number of considerations. These included the need to
• focus on those information needs that contributed to the goals and objectives of
the organization;
• capture the information needs of the diverse functions within the organization;
• provide sufficient detail on which to base a future information and IT strategy;
3
• implement a strategy that could be applied by a group of 10 project team
members, only one of whom had any background in information management;
• enable transfer of knowledge from the consultants to the project team members so
that they could maintain and continually develop the information architecture;
• gain participation and support of staff in the organisation as a whole, (staff were
already jaded by previous organization-wide surveys);
• complete the project in a short time frame i.e. three months.
As can be seen from Figure 1, Stage 1, ‘Project orientation’, relates to project start-up
which involved the identification of relevant orientation and background information,
familarisation with the corporate plan, tour of key areas in the organisation, briefing
sessions and the establishment of communication channels and protocols to facilitate
project work. Various IT tools, databases and applications used in the organisation, as
well as the IT proficiency of staff and the project team members were noted. This
helped to determine the platform for the database application and the subsequent
electronic survey.
Stage 2, ‘Task analysis and information needs analysis of staff’, signals the
commencement of the data collection process design and actual data collection. Key
areas that directly contribute to the organization’s mission and thrusts were identified
throughout all the Divisions. Staff who played key roles in these areas were identified
to participate in the exercise. A task analysis and information needs form was
designed (by the consultants) for completion by these key staff. The completion of
these forms was facilitated by project team members. These forms were used as the
basis for the design and implementation of the TA-IN database. A related set of
database input forms were designed for the project team members to input data into
the database. Finally, a number of queries were formulated to yield answers to a host
of typical information-related questions. Following a thorough pilot test on the
database, a final TA-IN database was generated to accept real user responses.
4
Start
Stage 1
Identify IT tools/databases/applications
Project Orientation
Stage 3
Task Analysis and
Information Needs
Analysis of Directors
Stage 5
Observations and
Clarification Exercise
Stage 8
Data Analysis: Task
Analysis and
Execute queries to generate findings;
Information Needs
Modify existing queries/generate new quries to support new
Database
analysis requirements;
Export database data to external applications (e.g. spreadsheet)
Stage 9 for further analysis and data presentation;
Data Analysis:
Information Audit
Database
End
5
Data collection using an array of techniques including semi-structured interviews with
directors, task analysis and focus groups with divisional staff, observation and
clarification sessions sought to progressively build up a rich picture of the information
needs of the organization. Data collected at these stages (3 through 5) was
incrementally added to the database by project team members using the database input
form. Feedback was solicited at these various stages to fine-tune the input form and
database design. This helped to improve the ease of data entry, promote the use of
standard vocabulary to describe similar information needs and sources, and enhance
the data integrity in the database (see Section 3.2). In order to distinguish the data
collected at these various stages, the input information was tagged with different
unique identifiers.
With all the data collected and stored in the two databases, Stages 8 and 9 were
primarily concerned with data analysis. The formulated queries were executed to
yield findings in various forms and formats. Existing queries were modified while
new queries were added as the iterative process of data analysis proceeded. Subsets
of data from the database were also exported to other applications, such as
spreadsheets, for further analysis and data presentation. Some of these sample outputs
are shown in Section 4 of the paper.
6
3. Database System Design and Implementation
Two databases, namely, TA-IN database and the IA database, were designed to
handle the data storage and analysis of the data derived from the qualitative and
quantitative approaches used in this project. The TA-IN database was used to
capture the various tasks, sub tasks and the type of information required to achieve
these tasks. It was also used to fully record additional information that was elicited or
volunteered by staff during the data collection process. In contrast, the IA database
was structured to capture the respondent's feedback on the categories of information
in terms of their perceived value, need/use and availability.
member_of
uses/
requires uses/
requires
Media
Organization_Data InfoType
Subject
InfoFlow Form
Centre Division Group
Coverage
InfoNeedGen
Info_Data
Frequency
InfoStatus
Quantity
InfoCode Whom
ModeOfDelivery OrgNameProduct
Figure 2. Entity relationship diagram for Task Analysis and Information Needs (TA-
IN) Model
Figure 2 shows the overall database model used for the TA-IN database using an
entity relationship diagram (or data object diagram). The entity types are represented
by boxes, their attributes by ovals, and the relationships between entities by
diamonds. Connections between entities are shown using standard symbols to
indicate cardinality and modality (Tillman, 1993). As shown in the data model,
every staff (Staff_Data) is a member within the hierarchical structure of Department,
7
Division or Group in the organization (Organisation_Data). Staff accomplish a series
of tasks (Task_Data) in their work. This work can be associated with daily role-
related routines, programmes or projects. Information is needed or generated in the
process of accomplishing the tasks. Main tasks can be broken down into sub tasks.
Such a child-parent relationship is uniquely identified by an extensible Task ID
identifier in the data model. For instance, MSD100, MSD100.1, MSD100.2,
MSD100.3 denotes a main task (labeled as 100) that comprises three sub tasks
(labeled as 100.1, 100.2 and 100.3) in the MSD department. Such a convention can
obviously be extended further to include sub-sub tasks if necessary. The information
needs and information generated are represented by the Info_Data entity. Through its
attributes, information can be defined exhaustively accordingly to the needs of the
information architecture project. In our example, information can be flowing in or
out (denoting information needed/used, or information generated), internal or external
(within or outside the organisation), for/from whom the information is
generated/obtained, information subject matter, category, frequency, mode of
delivery, media, and so on.
8
Table 1 shows a brief description of the data attributes in the model. It should be
noted that there is also a direct connection between the Staff_Data and Info_Data
entity. This is to capture information needs in instances where the information is not
directly related to particular tasks. For example, there may a common need among
Divisional Directors for information on benchmarking and best practices on
equivalent organizations abroad as part of their conceptual or current awareness
requirements.
InfoDesc
Title FirstName LastName
InfoCode InfoUse
uses/
Staff_Data Info_Topic
requires
InfoStatus InfoValue
member_of
InfoFrequency
Organization_Data
Figure 3. Entity relationship diagram for Information Needs Survey (IA) Model
Figure 3 shows the overall database model used for the IA database. Similarities exist
between the two databases. As before, every staff is a member of the organisation so
that the Staff_Data and Organisation_Data entities are identical. Every staff has his or
her own set of information needs that is defined in the Info_Topic entity. This entity
is related to TA-IN Info_Data entity in the form of the information code and
categories. Thus, this common link allows data from both qualitative and quantitative
analysis to be complemented and contrasted in the data analysis stages. In addition,
the Info_Topic entity allows users to identify information categories used, its
frequency, its status (whether the information is available or not) as well as the value
of the information identified. Users were allowed to add in new information
categories by defining a new information code. Table 2 shows a brief description of
these new data attributes in the model.
Both TA-IN and IA database were developed using the MS Access 97 (Microsoft,
1999a) platform according to the data models outlined in the previous section. This
was in line with the preferred type of database system in the organization. In
addition, all the project team members had this application installed on their
computers thereby enabling a convenient and de-centralized means to enter the
9
information needs data captured during the qualitative data collection phase of the
project.
Data entry to the TA-IN database was accomplished through a series of forms for the
convenience of the project team members and to minimize erroneous entries. The
select-and-click paradigm using pull-down menus was used whenever possible. A set
of example data entries showing different typical information needs/sources was
prepared beforehand and distributed as part of the ‘training session’ to promote
standardization in describing the same information entities.
Likewise, the electronic information needs survey was accomplished through two sets
of forms in MS Access 97 (Microsoft, 1999a) and MS Excel 97 (Microsoft, 1999b)
format to cater for different staff preferences and computing facilities of PSB staff.
As the organization-wide survey encompasses all professional staff with differing
levels of digital competency, due attention was placed on the design of these survey
forms to minimize ambiguity, and to make it as easy to complete and return as
possible. In addition, the overall process of the electronic survey followed a well-
defined procedure of survey preparation, distribution, data collection and processing.
Interested readers can refer to a separate paper detailing the implementation of the
electronic survey process (Foo & Hepworth, 1999). The paper also contains a section
of practical, tested guidelines (derived from the experience in this project) that serve
as a useful guide for similar work. As an example of these forms, Figures 4 and 5
shows a partially completed information needs questionnaire in MS Access and Excel
respectively.
10
Figure 4. Partially completed information audit questionnaire in MS Access
It should become apparent that in these survey forms, all unnecessary features of the
electronic form (e.g. database or spreadsheet menus, macros, short-cuts, etc.) are
disabled to ensure that staff are only presented with the bare minimum (but sufficient)
functionality to complete the survey. This is to minimize the temptation for users to
explore and create problems for themselves and the project team. Furthermore, in this
11
case where multiple forms were required, an attempt was made to create a similar
‘look-and-feel’ survey to achieve consistency. Clear instructions for survey
completion, appropriate use of language and tone, minimizing user input through the
use of list boxes for user selection, providing help features and following a logical
organization of questions were carefully observed and implemented in the survey
questionnaire.
The notion of ‘every field in the database is queryable’ is important so that maximum
use can be made of the databases for data repository and analysis. Once the data
model is properly linked and implemented through the appropriate fields and tables in
the database, the querying supported by databases becomes the value-added facility to
allow the captured information to be filtered, sorted, extracted, manipulated and
presented according to user needs. In the TA-IN and IA databases, a total of
approximately 30 basic queries were formulated. This number is in fact higher since
the analyst can apply slight modifications to the existing queries or input the desired
properties to change the focus of desired information. For example, a basic query can
be formulated to reveal the type of information sources used. This can then be filtered
to yield the information sources used by an individual staff, by role, by department,
by division or by group. An append query was also written to append the electronic
survey returns directly into the IA database. In this case, only the needed information
categories that are indicated by staff are appended. The queries were formulated
using the Access Query-By-Example (QBE) facility or direct SQL (Structured Query
Language) statements.
The query results coupled with data analysis helped to address many questions. These
included:
12
4. Information Profile Derived from Information Needs Analysis
In addition to using tabulated outputs or listings from the queries, many other forms
of result presentations are possible to fully illustrate the information profile derived
from such an information needs analysis. This section provides a number of examples
to show some of these possibilities. Mock up data and scaled-down versions of the
data are used in these figures.
80
60
40
20
0
)
)
file
)
ct)
y p s)
er
ive )
)
t)
try
e m ct)
- b uct)
rt)
tac e)
s)
nt)
m pab )
pe ip w er)
)
ng
pa - list
r)
)
)
B
ke
Co mpa t - b ogy
hn on)
tio n ..
)
d
s
ion ncy
ny lities
tis
ro
du
rd
..
NC
po
on ctur
ipa p...
te
itte spe
us
PS
ce
ar
SB h...
ve
br s...
na f...
b
e.
ini
...
ar
- c c...
lis rod
er
(p
or
io
em
i
ro
nd
t re
ind
l
- te ficat
Co ame
pa al s
(lis pro
Co nsa on i
ipa ulta
-e
o
ra
Co list - E, M
ym
ar stry
dv
u
ob
i
o
xp
tic ipat
ith
wi
ny
C2 7. C ny ( chn
p
r
pr
-t
(ca l st
(a
n
ic
by
r/e
yj
ti
s
w
y
ti
pa tion
-s
t
ar
u
n
b
rtic
n
ar con
er
M
n
te
a
ga ns b
C3 omp y ne
ind
rte
ipa ince
C0 any
tio
t-
m
(lis list -
-c
ipa
C2 any PS
t
om nsh
pa
by
c
c
tic
P
t
pa atio
pa
ll/c
- li
po
s
m
-
pa ene n -
C2 titor
Co ny (
b
o
i
p
on
(
n
ny
pa al's
,
-
m
rs
im
o
ici
E
om
tic
ny
-
tio
ny
Co Com nis
fp
o
2.
ar
on
pa tion
o
i
a
Co
i
Co - SM
co
ist
lat
e
C1 (pa ipat
i
i
pa part
a
tic
Co ny ( ipat
i
r
u
ipa
(p
t-
o
pa
t
(
fp
t
ny
(c
a
i
C2 y (l
at
re
Co ivid
(
m
C1 on
(p
an (par
C2 mp
a
fit
3.
ny
ny
o
tic
m
ny
p
Co artic
(
C0 C05 (or
y(
Co ticip
ic
tic
n
st
pa rtici
ny
ny
o
d
Co ting
a
pa (par
pa
(c
C
p
pa
ny
pa
Co pan
C2 y (li
C0 y (in
ar
at
ny
m
ny
pa
9.
r
p
9.
ny
b
m
m
pa
a
m
(d
m
(p
is
0.
C0 ny (
6.
y(
pa
Co y (p
n
Co
m
m
Co any
pa
l
Co
m
y(
pa
ny
ny
.
m
pa
C2 Co
m
ny
5.
Co
pa
m
n
0.
m
7.
C1 . Co
p
8.
1.
C1 mp
m
pa
pa
C2
m
9.
Co
4.
1.
C0
1.
Co
m
m
Co
C2
4.
Co
5
C3
2.
C
4.
Co
Co
C1
C0
6.
2.
8.
C2
7.
6.
1.
8.
C1
0.
3.
C0
3.
C1
C0
C1
C1
Figure 6. Information landscape showing demand for categories of company information across the whole
organisation
Figure 6 shows the information landscape that defines the demand for categories of
company information across the whole organisation. This can be extended easily to
incorporate all the information categories. Peaks and troughs are easily identifiable
from this plot. Figure 7 uses column bars to show the demand for company
information by division. This provides a means to compare the contrasting need for
information by different divisions. Figure 8 uses a doughnut plot to show the top five
information needs of the organization. Figure 9 uses a pie chart to show a similar set
of information in the form of the top 5 information sources.
13
13%
24%
C01
. Co
mpa
C03 ny (c
. Co omp
mpa
0
20
40
60
80
100
ny (l any
C02 nam
12%
istin . Co e - li
C04 g of
pers mpa st)
. Co ons ny (p
mpa rofile
ny (o by jo
C06
. Co C05 rgan b fu )
mpa . Co isati ncti
ny (i mpa ona on)
ndiv ny (c l str
idua
l's p a ll/con
uctu
re)
C07 artic tact
C08 . Co ipati repo
. Co mpa on in rt)
%
mpa ny (p PSB
ny (p artic ac
C09 artic ipati
C10
. Co . Co ipati o n-e
mpa on - ven
mpa
ny (p ny (p con t)
C11 artic artic sult
. Co ipati anc
mpa ipati on - y)
on -
C12 ny (p
artic in cen
train
C13 . Co
mpa ipatio tiv e sc
ing)
. Co ny (p n - li hem
mpa brary e)
ny (p artic
artic ipati sch
on - eme
C14 ipati cert )
. Co
mpa
on -
te ificatio
chn
n ic n)
24%
C16 C15 y a
. Co . Co (part l se
mpa mpa icip rvic
C17 ny (p ny (p atio es)
27%
. Co artic artic n-a
mpa ipati ipati ward
ny (b on - on - s)
C18 ene indu stan
. Co fit o stry dard
mpa f pa prog )
14
ny (d rticip ram
ata a tion me)
on tr in P
ans SB
C19
. Co a c tion a c ti
C20 mpa with
. Co ny (P PSB
C21 mpa
ny (c SB )
omp with
etito PSB
C24 C23 r-c )
. Co apa
mpa . Co
m pan b ilities
C25
ny (l
ist - y (a )
. Co SME dve
rtise
mpa
n
, PS
M r)
C26 y (list E, M
. Co - by NC)
mpa tech
C27 ny (l nolo
C28 . Co ist - gy)
. Co mpa by p
mpa ny (l roce
ny (l ist - ss)
ist - by p
C29 rodu
. Co by n
ew ct)
C30 mpa prod
ny (l
C31 . Co
mpa is t - by
uct)
. Co ny (l mark
mpa
ny (l ist - et)
ist - by in
imp dus
orte try)
r/ex
port
er)
IND
Kompass
12%
RCB
25%
Business Times
18%
Dialog
23%
Company annual
reports
22%
Figure 10. Table plot showing common information needs or gaps relating to
company information categories by division
15
LED LED
P10, OR08
C32, I07, SU23
P10
AFD AFD
C02, C31
P14, P15
SMD SMD
P10
IND IND
Information Information
In Out
Division Division Division
Figure 11. Information flow diagram showing internal information interaction from
one division with itself and others
The table plot (2D matrix) of Figure 10 is used to show the common information
needs or gaps relating to the company information category by division. Figure 11
uses a data flow diagram to show the internal information flow between a division and
all other divisions in the organization. Information flowing in to the division from
other divisions is shown on the left while information flowing out to other divisions is
shown on the right. Futhermore, internal information generated and consumed by the
same division is shown by a loop in the figure. A similar figure can be used to map
the information flow of the organization with the external environment.
16
Figure 12. External customized application to interface with IA database
5. Conclusion
This paper has traced the design and development of two databases that served as an
accompanying tool to carry out an information architecture project for a large
organisation of approximately 1000 people. The TA-IN and IA databases were used
as a repository for data collected during the qualitative and quantitative phases of the
project. Input forms provided a convenient means to input the qualitative data, while
an append query function provided an efficient way to append the relevant data from
the electronic survey directly into the IA database. These databases subsequently
served as complementary tools during data analysis by allowing various queries on
the information to be performed. It also allowed the easy export of data to other
applications so that other more detailed analysis could be carried out. These
databases are now available for subsequent analysis and also to facilitate the capturing
and analysis of additional information needs in future. Without these two databases it
would have been very difficult to manage the large amount of data gathered through
the use of both qualitative and quantitative methods.
The careful design and implementation of the methodology, data collection and
analysis has led to the completion of the information architecture project in three
months, which for a project of this scale, is extremely short. Internal PSB documents
and business process re-engineering consultants had estimated between six months
and one year for the project. The project laid the groundwork and generated findings
17
for the derivation of PSB’s strategic information plan. This plan will, it is expected,
enable the realization of the corporate vision to become a knowledge-based enterprise
where the systematic management of data, information and knowledge is seen as a
key strategy for success.
References
Wilson, T. D. (1994). ‘Information Needs and Uses’, In Vickery, B. (ed). Fifty years
of information progress: A Journal of Documentation Review. London: Aslib
15-51.
Wixen, D. Ramey, J. (1996). Field methods casebook for software design. New
York: Wiley Computer Publishing
18